GitHub - vidraj/npfl104-aspect: Detect aspect of Czech verbs: a machine learning task for NPFL104 2016/17.

vidraj / npfl104-aspect Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Detect aspect of Czech verbs: a machine learning task for NPFL104 2016/17.

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 26 Commits
.gitignore		.gitignore
Makefile		Makefile
README		README
draw-correlation.py		draw-correlation.py
get-features.py		get-features.py
get-seeded-random.sh		get-seeded-random.sh

Repository files navigation

# Aspect

Predict the aspect of verbs.

Format:
	Features are binary values specifying whether an n-gram described by the feature occurs in the verb.
	Output value (the last boolean on each line) specifies aspect: 0 == perfective, 1 == imperfective.

Names of the verbs (the data labels) are in {test,train}-labels.txt
Header describing the n-grams is in header.txt
To get the complete dataset with labels and the header in place, run
$ cp header.txt complete-data.txt && paste -d, {test,train}-labels.txt {test,train}.txt >> complete-data.txt